Metabolic consequences of cystinuria

Cystinuria is one of the most common autosomal recessive genetic disorders with an incidence of 1 in 7000 worldwide [1]. Patients suffer from significant morbidity and few new options have been developed in the past 20 years [2]. Cystinuria is caused by an inherited defect in the transport of cystine and dibasic amino acids (ornithine, lysine, and arginine) in renal tubular cells. Cystine is not soluble in urine so kidney and bladder stones form when the renal tubule fails to reabsorb the amino acid back into the bloodstream. Recurrent cystine kidney stones are associated with pain, frequent urinary tract infections, bleeding, urinary tract obstruction, need for multiple surgical procedures, and kidney failure [2]. Medical treatment to prevent the formation of cystine stones is not very effective and has many unpleasant side effects [3]. Mutations in the gene encoding a renal amino acid transporter (SLC3A1, or the rBAT protein) have been identified in most patients with cystinuria establishing the most common cause of the disease termed cystinuria type I [4]. A second gene SLC7A9, encoding for b^0,+type amino acid transporter 1 (b^0,+AT) accounts for a smaller fraction of cases [5, 6]. The rBAT protein (Slc3a1 gene product) heterodimerizes with the b^0,+AT protein (Slc7a9 gene product) within proximal tubular cells in the nephron, thereby mediating resorption of cystine and dibasic amino acids from the urine [7]. Knowledge of the genetic basis of cystinuria provides opportunities for developing treatments.

Mouse models of cystinuria type I have been generated. Peters et al. identified a missense mutation in Slc3a1 in a N-ethyl-N-nitrosourea (ENU) mutagenesis screen in C3HeB/FeJ mice [8]. The authors demonstrated reduced kidney weight in homozygous males as compared to wild type males and homozygous females. Additionally, plasma urea levels were elevated in homozygous males from 13 to 20 weeks old as compared to wild type animals. Livrozet et al. identified a spontaneous mutation in Slc3a1 in 129SvPasCrl mice resulting in cystinuria [9]. Renal function was not significantly impaired in the mutant mice as measured by serum creatinine measurements. However, they did observe increased macrophages and interstitial fibrosis in the mutant mice. Ercolani et al. demonstrated bladder outlet obstruction in male cystinuria mice on a mixed C57Bl/6 and 129/SvJ backround [10]. The genetic strategy for generation of this cystinuria mouse line has been reported [11]; however, we sought to more fully characterize this cystinuria mouse strain.

Recently, it has been debated whether or not kidney stone formation is a pathogenic process contributing to kidney disease [12]. Blockages produced by crystals of various composition have been reported to lead to chronic kidney disease, acute kidney injury, renal colic, or nephrocalcinosis [13]. Crystal deposits may be found in the nephron, renal vasculature, and/or urinary tract [13]. Beyond direct blockage, additional molecular mechanisms have been explored including involvement of the inflammasome and crystal granuloma formation [14, 15]. It has been observed that cystinuric patients develop chronic kidney disease (CKD) even more commonly than usual stone formers [16‐18]. Hypertension has been shown to be associated with CKD in cystinuria patients [19]. Cystinuria has been associated with lower renal function and quality of life compared to other kidney stone formers [20]. Nonetheless, little is known about what contributes to the development of CKD in patients with cystinuria other than stone formation. We sought to more fully characterize the cystinuria mouse strain previously reported by Ercolani et al. by evaluating plasma amino acid levels and to determine if this strain develops CKD over time [10]. Additionally, though one previous report found decreased glutathione levels in leukocytes of patients with cystinuria [21], glutathione levels have not been studied in other tissues or in Slc3a1^−/− animals.

We first confirmed knockout of Slc3a1 using RT-PCR to evaluate expression levels in various tissues. As expected, Slc3a1 RNA levels were dramatically reduced in tissues where expression normally occurs, such as the kidney and small intestine (Fig. 1a). Additionally, as expected, Slc7a9 RNA levels were unaltered by knockout of Slc3a1 (Fig. 1b). We next used Western blot and immunofluorescent microscopy to evaluate for loss of rBAT expression in kidneys of Slc3a1^−/− mice. Western blot demonstrated loss of rBAT expression from kidney lysates (Fig. 1c). Immunofluorescent microscopy for rBAT demonstrated that knockout of Slc3a1 resulted in loss of rBAT expression in the proximal tubules of Slc3a1^−/− mice when compared the wild type mice (Fig. 1d and e).

Knockout mice did not develop kidney stones but did develop bladder stones detectable by x-ray as reported previously (Fig. 2a) [10]. Consistent with previous observations, we observed gender differences between male and female Slc3a1^−/− mice with female mice very rarely developing bladder stones [10]. We did not observe a high rate of bladder stone formation in male Slc3a1^−/− mice on normal chow (0.31% cystine) (Table 2). However, when we placed the male Slc3a1^−/− mice on breeder chow (0.36% cystine), we observed a higher rate of bladder stone formation. Comparing the % of mice with bladder stones at 28 weeks, 100% of Slc3a1^−/− mice on breeder chow had bladder stones whereas only 42% of Slc3a1^−/− mice on regular chow exhibited bladder stone formation (Fig. 2b). The rate of bladder stone formation in mice on the breeder chow is depicted in Fig. 2c. These results demonstrate that dietary intake can have a major impact on the rate of stone formation in Slc3a1^−/− mice.

We observed that male Slc3a1^−/− mice were consistently lower in weight when compared to Slc3a1^+/+ mice of the same age (Fig. 3a,b). This weight difference was not recapitulated in the female mice. All mice exhibited the same length from their nose to the tip of their tail, indicating an overall metabolic phenotype rather than nutritional deficiency contributed to the difference in weights of the male mice raised and maintained on breeder chow (Fig. 3c). We evaluated the plasma amino acid levels in Slc3a1^−/− mice and compared them to wild type mice (Table 3). Consistent with the reduced weight in male Slc3a1^−/− mice, we observed a more severe phenotype with multiple amino acid levels different between wild type and knockout animals. Evaluation of amino acids in both male and female Slc3a1^−/− mice revealed only ornithine, lysine, and taurine to be reduced in both sexes. Therefore, knockout of rBAT resulted in reduce plasma ornithine and lysine in these animals. Cysteine, which forms cystine when two molecules are joined together, is metabolized to taurine [24] which was also reduced across sexes. Interestingly, plasma cystine was reduced only in male Slc3a1^−/− despite both male and female mice exhibiting cystinuria [11].

Previous reports showed elevation of BUN in some Slc3a1^−/− mice strains but no elevation in serum creatinine in others [8, 9]. We therefore sought to evaluate for the development of CKD in this mouse strain with a mixed C57Bl/6 and 129/SvJ background [10]. Interestingly, male mice fed only normal chow showed elevated BUN, whereas male mice fed breeder chow showed both elevated BUN and creatinine (Fig. 4). Mice evaluated were between 40 and 87 weeks old (Additional files 1 and 2). Histologic evaluation of the kidney tissue from Slc3a1^−/− mice with higher creatinine on normal chow (Fig. 5b) demonstrated dilated tubules from obstruction and increased fibrosis when compared to those with normal creatinine (Fig. 5a). The observed dilation of tubules and fibrosis was also observed in mice with elevated creatinine (Fig. 5d) compared to normal creatinine when placed on the breeder chow (Fig. 5c). Histologic evaluation of the bladders of male Slc3a1^−/− mice revealed increased inflammatory infiltrate within the bladder wall and cystine stones in the bladder in mice having a normal serum creatinine (Fig. 5f) when compared to the bladders of wild type mice (Fig. 5e). Some Slc3a1^−/− mice with an elevated serum creatinine showed embedded cystine crystals within the bladder wall in addition to bladder inflammatory infiltrate (Fig. 5g).It is known that cystinuria results in stone formation leading to obstruction and CKD. However, we sought to determine if there might be other metabolic consequences of cystinuria beyond just stone formation. Glutathione, or γ-l-glutamyl-l-cysteinyl-glycine (GSH), serves as a detoxifier, protector from oxidative damage, and is arguably the most important low molecular weight antioxidant in cells [25]. The availability of cysteine, formed by two molecules of cystine, is a major determinant of the regulation of GSH synthesis [26]. Therefore, cystine deficiency could affect GSH levels. Martensson et al. previously reported decreased leukocyte GSH levels in homozygous cystinuric patients [21]. We sought to evaluate GSH levels in male Slc3a1^−/− mice which exhibited lower plasma cystine levels. Therefore, we evaluated the levels of reduced (GSH) and oxidized (GSSG) glutathione in the livers of Slc3a1^−/− mice. We observed lower GSH and lower GSSG in the livers of Slc3a1^−/− mice compared to WT controls (Fig. 6). The total amount of glutathione (GSH + GSSG) was lower (Fig. 6), implying a possible connection between the lower availability of cysteine and total glutathione reserve. A reduction of the ratio of reduced to oxidized forms of glutathione was also observed (Fig. 6), suggesting that the ability of the liver to respond to oxidative stress is reduced in Slc3a1^−/− mice.

Treatment for cystinuria has not dramatically changed in the past 20 years despite greater understanding of the genetic basis [1]. Mouse models for cystinuria not only confirm the genetic basis of the disease but also provide models for testing and evaluating new therapies [10]. There are likely genes that affect stone formation in patients with cystinuria that could be derived from analysis of patients and then tested in mouse models. There are likely genes that affect stone formation in patients with cystinuria that could be derived from analysis of patients and then tested in mouse models. Recently, Zee et al. demonstrated that α-lipoic acid prevents cystine stone formation in the mouse model reported in this article [11].

Three type A cystinuria mouse models have been generated including the one described herein that was generated by exon 1 deletion [10]. The other models consist of spontaneous mutation (E383K) [9] or ENU induced mutagenesis (D140G) [8]. The D140G model on a 129S2/SvPasCrl background demonstrated a 40% reduction in survival by 15 weeks with elevation in serum BUN though no increase in serum creatinine or interstitial kidney fibrosis [9]. Possible gender differences in these observations were not commented on though stones mainly occurred in males [9]. The E383K model on the C3HeB/FeJ background showed elevated serum BUN in 20 week old male mice and decreased kidney weight in 32–48 week old male mice with neither phenotype observed in female mice [8]. The observed gender difference in stone formation between male and female in the context of cystinuria has been observed in patients and reviewed elsewhere [27, 28]. The difference in stone formation does not appear to be due to differences in urinary cystine levels, but it may due to differential cystine aggregation in male versus female urine [27, 29]. Nonetheless, our study demonstrates that CKD does occur in aged male Slc3a1^−/− mice.

Little is known about what factors affect the phenotype of cystinuria and progression to CKD. [2] We found that the diet fed to the Slc3a1^−/− mice had a dramatic effect on the rate of stone formation. Common treatment for cystinuria includes limiting dietary sodium and protein intake, increased fluid intake, urinary alkalization, and possibly even thiol drugs and captopril [1]. A previous study demonstrated that a low protein diet with more plant protein sources reduced cystine excretion in cystinuria patients [30]. We have not found any studies on the role of diet in cystine stone formation in mouse models of cystinuria. Human clinical studies on the effects of diet for potential modification of human cystinuria are lacking [2]. Future studies could be directed at evaluating dietary changes in a more controlled manner to tease out modifications which could slow the progression of stone formation in cystinuria.

The inability to reclaim cystine from the urine likely alters cystine metabolism in the whole animal. Knockout of Slc3a1 affects its expression not only in the kidney but also the intestine. The intestinal peptide transporter Pept1 (Slc15a1) is thought to permit reabsorption of cystine and other amino acids from the gut somewhat compensating for loss of Slc3a1 in the intestine [31, 32]. How this might contribute to overall gender difference in cystinuria and our observed differences in plasma amino acid levels is currently unknown.

Our results reveal the metabolic phenotype to be more severe in male knockout animals when compared to females. Glutathione levels were lower in the Slc3a1^−/− mice. Given the importance of glutathione in detoxification and oxidative stress, cystinuria may result in an overall metabolic state of increased sensitivity to injury from a variety of mechanisms [25]. As patients with cystinuria have been reported to develop more CKD than other stone formers, lower levels of GSH could increase the sensitivity of Slc3a1^−/− mice to development of and recovery from injury resulting in worsened CKD. Previous research has demonstrated that anti-oxidant cystine metabolites, GSH, and taurine exert protection against progression of renal fibrosis [33‐35]. Additionally, others have found these to be protective against renal ischemia reperfusion injury [36‐39]. In patients, cystinuria results in kidney stones and obstruction. Reduced glutathione could result in worsened injury or inability to recover from injury in cystinuria when compared to stone formers with non-cystinuric disease pathology. Importantly, new therapies such as α-lipoic acid may reduce stone formation by increasing solubility of cystine in the urine but have no effect on regaining cystine transport [11]. Therefore, such therapy may reduce stone formation but may not improve cystine or oxidative metabolism overall. Our data suggest that diet manipulation, supplementation, and/or drugs that work to improve the ratio of reduced to oxidized glutathione in vivo should be tested in mouse models for their effect on the development of CKD. Such studies would provide further guidance for care of cystinuric patients and hopefully guide the development of new therapies to help prevent kidney disease progression in this population.

We have demonstrated that dietary intake can have an impact on the severity of the phenotype of cystinuria both in regards to stone formation and the development of CKD in a mouse model of cystinuria. Male mice exhibited a more severe phenotype concerning animal weight and plasma amino acid composition. Cystinuria affected glutathione composition in the liver. Future studies will systematically evaluate dietary components on the cystinuria phenotype and what effect the resultant alteration in glutathione metabolism has on physiology and pathophysiology.